Post-traumatic Stress Disorder in Older U.S. Military Veterans: Prevalence, Characteristics, and Psychiatric and Functional Burden

Abstract

Objective:

To characterize the prevalence, characteristics, and comorbidities of subthreshold and full post-traumatic stress disorder (PTSD) in older U.S. military veterans.

Design and Setting:

A nationally representative web-based survey of older U.S. military veterans who participated in the National Health and Resilience in Veterans Study (NHRVS) between November 18, 2019 and March 8, 2020.

Participants:

U.S. veterans aged 60 and older (n = 3,001; mean age = 73.2, SD: 7.9, range: 60–99).

Measurements:

PTSD was assessed using the PTSD Checklist for DSM-5. Self-report measures assessed sociodemographic characteristics, trauma exposures, suicidal behaviors, psychiatric and substance use disorders, as well as mental, cognitive, and physical functioning. Multivariable analyses examined correlates of subthreshold and full PTSD.

Results:

The vast majority of the sample (n = 2,821; 92.7%) reported exposure to one or more potentially traumatic events. Of those exposed to such events, 262 (9.6%, 95% confidence interval [CI]: 8.4%–10.9%) and 68 (1.9%, 95% CI: 1.3%–2.6%) screened positive for subthreshold and full PTSD, respectively. The prevalence of subthreshold and full PTSD was significantly higher in female veterans and veterans who use VA as their primary healthcare. Subthreshold and full PTSD groups endorsed more adverse childhood experiences and total traumas than the no/minimal PTSD symptom group, the most common traumatic experiences endorsed were combat exposure, physical or sexual assault, and life-threatening illness or injury. Veterans with subthreshold and full PTSD were also more likely to screen positive for depression, substance use disorders, suicide attempts, nonsuicidal self-injury, and suicidal ideation, and reported lower mental, cognitive, and physical functioning.

Conclusion:

Subthreshold PTSD and full PTSD are prevalent and associated with substantial clinical burden in older U.S. veterans. Results underscore the importance of assessing both subthreshold and full PTSD in this population.

OBJECTIVE

Although many adults will be exposed to a potentially traumatic event by the time they reach older adulthood,¹ only a subset develop posttraumatic stress disorder (PTSD). Lifetime prevalence estimates of PTSD range from 4% to 7%^2,3 and current prevalence estimates range from 1% to 3%^4–6 but are higher in older women,^3,7 those who experience more trauma,⁸ and those with combat trauma.⁹ In older adults, PTSD is associated with worse cognitive,¹⁰ physical,¹¹ and social¹² functioning, and is associated with elevated rates of mood, anxiety, and substance use disorders, as well as increased risk for suicidal ideation and attempts.^13–16

Rates of subthreshold PTSD in older adults are higher than those observed for full PTSD and range from 7% to 13%.^4,5,17–19 Subthreshold PTSD, which is characterized by clinically significant PTSD symptoms below the threshold for a diagnosis, is not a recognized diagnostic entity and is less well studied in older adults despite its higher prevalence and potential clinical significance. For example, a nationally representative study of 9,463 U.S. adults aged 60 and older found that lifetime subthreshold PTSD was slightly more prevalent than PTSD (5.5% versus 4.5%), chronic in nature, and associated with elevated rates of co-occurring psychiatric disorders and reduced psychosocial functioning³. Subthreshold PTSD in older adults may represent an emergence or re-emergence of PTSD symptoms later in life as older adults confront aging-related stressors²⁰ such as illness, bereavement, role transitions, and cognitive changes. Older adults with PTSD or subthreshold PTSD may present with more engagement with traumatic memories and less avoidance,²¹ consistent with the theory that some older adults may re-engage with earlier life experiences in an effort to form meaning or understanding as part of life review processes in older age.²²

Such considerations may be particularly important for older military veterans who have high rates of trauma exposure and may experience re-emergent PTSD later in life.²² Currently, nearly half (46%) of U. S. veterans are age 65 or older,²³ most of whom are aging Vietnam-era veterans who may be at elevated risk for PTSD.²⁴ Rates of PTSD in older U.S. military veterans are typically higher than those observed in the civilian population, varying between 1% and 22% with a pooled prevalence of 8% in a recent meta-analysis²⁵; rates of PTSD are generally lower in older veterans as compared to younger veterans.²⁶ While large-scale surveys in Germany,⁴ the Netherlands,⁶ and Australia²⁷ have described the prevalence of subthreshold PTSD and PTSD in older civilian populations in these countries, no known study has examined the prevalence, characteristics, and clinical burden of PTSD and subthreshold PTSD in a nationally representative sample of older U.S. military veterans. To address this gap, we analyzed data from a contemporary, nationally representative sample of older U.S. military veterans to evaluate the following three aims: 1) characterize the prevalence of PTSD and subthreshold PTSD; 2) identify sociodemographic, military, and trauma characteristics of veterans with PTSD and subthreshold PTSD; 3) evaluate clinical burden by examining psychiatric and functional correlates of PTSD and subthreshold PTSD.

METHODS

Sample

The National Health and Resilience in Veterans Study (NHRVS) was first launched in 2011 and has recruited three independent, prospective cohorts of U.S. military veterans. In the current study, we analyzed data from the most recent 2019–2020 cohort, which surveyed 4,069 U.S. veterans between November 18, 2019 and March 8, 2020 (median completion date: November 21, 2019); the current study focused on veterans aged 60 and older (n = 3,001; mean age: 73.2, SD: 7.9, range: 60–99). Veterans completed a 50-minute, anonymous, web-based survey. The NHRVS sample was drawn from KnowledgePanel, a research panel of more than 50,000 U.S. households maintained by Ipsos, a survey research firm. KnowledgePanel is a probability-based, online, nonvolunteer access survey panel of a nationally representative sample of U.S. veterans that covers approximately 98% of U.S. households. Panel members are recruited through national random samples, originally by telephone and now almost entirely by postal mail. KnowledgePanel recruitment uses dual sampling frames that include both listed and unlisted telephone numbers, telephone and non-telephone households, and cell-phone-only households, as well as households with and without Internet access. To permit generalizability of study results to the entire population of U.S. veterans, the Ipsos statistical team computed poststratification weights using the following benchmark distributions of U.S. military veterans from the most recent (August 2019) Current Veteran Population Supplemental Survey of the U.S. Census Bureau’s American Community Survey: age, gender, race/ethnicity, Census Region, metropolitan status, education, household income, branch of service, and years in service. An iterative proportional fitting (raking) procedure was used to produce the final poststratification weights. All participants provided informed consent and the study was approved by the Human Subjects Committee of the VA Connecticut Healthcare System.

Assessments

Demographics and participant characteristics

In addition to age, gender, race/ethnicity, education, marital status, employment status, and annual household income, we ascertained mode of enlistment in the military, combat veteran status (coded into 0, 1, or 2+ deployments), duration of military service (coded into <10 versus 10+ years military service), and whether the Veterans Health Administration is a primary source of healthcare.

Adverse childhood experiences

The Adverse Childhood Experiences (ACES) Questionnaire was used to assess for exposure to seven types of childhood maltreatment (e.g., physical abuse, emotional neglect), occurring between birth and age 18 years.²⁸ Items were summed for a total score, with higher scores indicating greater ACES.

Trauma exposure

The Life Events Checklist for DSM-5 (LEC-5) was used to assess exposure to 17 types of potentially traumatic events (PTEs).²⁹ For each PTE type, participants are asked to indicate their exposure type (whether the event “happened to me,” “witnessed it,” “learned about it happening to close family or friend” and/or “exposed to it as part of my job”). For analyses of traumas, disaster/accident was defined as endorsement of any of five items (natural disaster, fire or explosion, transportation accident, other serious accident, or exposure to toxic substance); interpersonal violence as any of four items (physical assault, assault with a weapon, sexual assault, or other unwanted sexual activity); combat or captivity with these two respective items; illness/injury as any of four items (life-threatening illness or injury, severe human suffering, sudden violent death, or sudden accidental death); and serious injury, harm, or death you caused to someone else was defined by a single item. Direct exposures were defined as endorsement of “happened to me” and indirect exposures were defined as any other exposure type. Cumulative direct trauma burden was calculated as the total number of direct trauma exposures to any trauma type (with a possible range from 0 to 17) and cumulative indirect trauma burden was calculated as the total number of indirect trauma exposures (with a possible range of 0–51). An additional question asked which of the endorsed PTEs “was the worst for you;” responses to this question were coded as the index trauma.

PTSD symptoms

The PTSD Checklist for DSM-5 (PCL-5) was used to assess PTSD symptoms in relation to veterans’ index trauma endorsed on the LEC-5.³⁰ The following two questions were added to assess Criterion F (duration) and G (functional significance): “How long did these reactions last?” and “Did these reactions cause you distress or result in a failure to fulfill obligations at home, work, or school?” Report of symptoms lasting for one month or longer was considered a positive endorsement of Criterion F. Endorsement of ‘moderate’ or greater distress or functional impairment was considered a positive endorsement of Criterion G. Following prior work,³¹ PCL-5 responses were used to create a three-group variable: a) no/minimal PTSD symptoms (defined as endorsement of ≤1 PTSD criteria B-E at a severity of “moderate” or higher); b) subthreshold DSM-5 PTSD (defined as endorsement of 2 or 3 B-E criteria, or all 4 B-E criteria but not 1 month symptom duration and/or functional impairment); and c) probable lifetime DSM-5 PTSD (defined as meeting criteria A-G for PTSD). Cronbach’s α on PCL-5 items in the current sample was excellent (0.94).

Mental health treatment

Lifetime mental health treatment was assessed with the question: “Have you ever received mental health treatment (e.g., prescription medication or psychotherapy) for a psychiatric or emotional problem?” Current mental health treatment was assessed with the questions: “Are you currently receiving psychotherapy or counseling for a psychiatric or emotional problem?” and “Are you currently taking prescription medication for a psychiatric or emotional problem?”

Psychiatric and functional comorbidities

Lifetime psychiatric and substance use disorders, as well as lifetime suicide attempt, past-year suicidal ideation and intent, and non-suicidal self-injury (NSSI) were measured as described in Table 3. Current mental, physical, and cognitive functioning were measured as described in Table 4.

TABLE 3.

Psychiatric and Mental Health Treatment Variables by Current PTSD Status in Older U.S. Military Veteran

	No/Minimal PTSD Symptoms N = 2,387 Weighted 88.5%	Subthreshold PTSD N = 262 Weighted 9.6%	Full PTSD N = 68 Weighted 1.9%	Bivariate Test of Difference (χ2)	Subthreshold PTSD Versus No/Minimal PTSD Symptoms OR (95% CI)	Full PTSD Versus No/Minimal PTSD Symptoms OR (95% CI)
Lifetime
Major depressive disordera	183 (7.1%)	77 (27.5%)	41 (62.5%)	207.08***	3.36 (2.29–4.95)***	12.60 (6.17−25.73)***
Alcohol use disordera	861 (36.9%)	149 (60.6%)	44 (63.4%)	52.72***	2.11 (1.54−2.88)***	2.15 (1.09−4.24)*
Drug use disordera	193 (8.4%)	57 (25.1%)	14 (17.5%)	57.76***	2.85 (1.92–4.23)***	1.34 (0.54–3.34)
Nicotine use disorderb	414 (19.0%)	68 (26.1%)	20 (32.5%)	9.87**	1.29 (0.91−1.94)	1.67 (0.83−3.38)
Non-suicidal self injuryc	20 (0.7%)	9 (3.4%)	5 (5.0%)	19.64***	3.42 (1.24−9.45)*	555 (1.11−27.68)*
Suicide attemptd	25 (0.9%)	20 (6.4%)	8 (10.0%)	55.11***	4.17 (1.84−9.43)**	5.26 (1.47−18.85)*
Current
Major depressive disordere	58 (2.7%)	46 (14.7%)	22 (37.5%)	162.01***	5.07 (3.04−8.48)***	18.24 (8.47−39.30)***
Generalized anxiety disordere	43 (2.1%)	23 (8.8%)	23 (30.0%)	118.01***	4.34 (2.32−8.13)***	26.75 (11.06−64.72)***
Alcohol use disorderf	167 (7.5%)	33 (13.8%)	11 (12.5%)	10.71**	1.82 (1.14−2.91)*	1.65 (0.62−4.35)
Drug use disorderg	136 (6.0%)	32 (13.4%)	11 (15.4%)	19.85***	1.77 (1.09−2.87)*	2.00 (0.78−5.14)
Suicidal ideationd	114 (4.4%)	50 (16.3%)	16 (20.0%)	61.71***	2.97 (1.86−4.73)***	3.50 (1.49−8.18)**
Suicidal intentd	8 (0.3%)	6 (2.0%)	4 (5.0%)	26.13***	3.97 (0.96−16.44)	6.90 (1.07−44.52)*
Lifetime mental health treatment	359 (13.2%)	120 (36.9%)	48 (72.5%)	166.12***	2.39 (1.69−3.38)***	8.95 (4.21−19.02)***
Current mental health treatment	131 (5.1%)	63 (19.7%)	27 (46.2%)	146.35***	2.96 (1.90−4.60)***	7.56 (3.71−15.43)***
Psychotropic medication	119 (4.7%)	56 (17.8%)	27 (45.0%)	143.85***	3.00 (1.90−4.72)***	8.36 (4.09−17.09)***
Counseling/psychotherapy	55 (2.0%)	37 (11.3%)	18 (35.0%	165.57***	3.59 (1.97−6.52)***	10.65 (4.71−24.11)***

Notes. OR: odds ratio; 95% CI: 95% confidence interval; Statistically significant association:

^*p < 0.05;

^**p < 0.01;

^***p < 0.001.

Bivariate tests of difference were χ² tests. Degrees of freedom for χ² tests = 2; and for each group comparison in regression models = 1. Odds ratios are adjusted for age, gender, race/ethnicity, marital status, enlistment/commissioned status, combat veteran status, number of deployments, VA healthcare status, adverse childhood experiences, and total lifetime trauma burden; analyses of non-suicidal self-injury, suicide attempt, suicidal ideation, suicidal intent, and mental health treatment are additionally adjusted for lifetime major depressive, alcohol, drug, and nicotine use disorders.

^aLifetime major depressive, alcohol, and drug use disorders (MDD, AUD, DUD) were assessed using modified self-report modules from the Mini Neuropsychiatric Interview for DSM-5. Sheehan DV, Lecrubier Y, Sheehan KH, et al. The Mini-International Neuropsychiatric Interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. The Journal of clinical psychiatry. 1998;59 Suppl 20:22–33;quiz 34–57.

^bLifetime nicotine use disorder (NUD) was assessed using the Fagerström Test for Nicotine Dependence, with scores ≥6 indicative of a positive screen. Heatherton TF, Kozlowski LT, Frecker RC, Fagerström KO. The Fagerström Test for Nicotine Dependence: a revision of the Fagerström Tolerance Questionnaire. Br J Addict. Sep 1991;86(9):1119–27. doi:10.1111/j.1360-0443.1991.tb01879.x

^cNon-suicidal self-injury (NSSI) was assessed using questions from the Self-Injurious Thoughts and Behavior Interview-Short Form. Nock MK, Holmberg EB, Photos VI, Michel BD. Self-Injurious Thoughts and Behaviors Interview: development, reliability, and validity in an adolescent sample. Psychol Assess. Sep 2007;19(3):309–17. doi:10.1037/1040-3590.19.3.309

^dLifetime suicide attempt (SA), past-year suicidal ideation, and suicidal intent were assessed using questions from The Suicide Behaviors Questionnaire-Revised. Osman A, Bagge CL, Gutierrez PM, Konick LC, Kopper BA, Barrios FX. The Suicidal Behaviors Questionnaire-Revised (SBQ-R): validation with clinical and nonclinical samples. Assessment. Dec 2001;8(4):443–54. doi:10.1177/107319110100800409

^eCurrent depressive (α = 0.82) and anxiety (α = 0.77) symptoms were assessed using the Patient Health Questionnaire-439, with scores ≥3 indicative of positive screens for depression and anxiety. Kroenke K, Spitzer RL, Williams JB, Löwe B. An ultra-brief screening scale for anxiety and depression: the PHQ-4. Psychosomatics. Nov-Dec 2009;50(6):613–21. doi:10.1176/appi.psy.50.6.613.

^fCurrent alcohol use disorder was assessed using the Alcohol Use Disorders Identification Test40, with scores ≥8 indicative of a positive screen (α=0.76). Bohn MJ, Babor TF, Kranzler HR. The Alcohol Use Disorders Identification Test (AUDIT): validation of a screening instrument for use in medical settings. J Stud Alcohol. Jul 1995;56(4):423–32. doi:10.15288/jsa.1995.56.423.

^gCurrent DUD was assessed using the Screen for Drug Use41, with a response of ≥7 days to the following question indicative of a positive screen: “How many days in the past year have you used non-prescription drugs?;” if the response to this question is 6 or fewer days, a response of ≥2 days to the question “How many days in the past 12 months have you used drugs more than you meant to?” is indicative of a positive screen for DUD. Tiet QQ, Leyva YE, Moos RH, Frayne SM, Osterberg L, Smith B. Screen of Drug Use: Diagnostic Accuracy of a New Brief Tool for Primary Care. JAMA internal medicine. Aug 2015;175(8):1371–7. doi:10.1001/jamainternmed.2015.2438.

TABLE 4.

Multivariable Analyses of Functioning Measures by Current PTSD Status in Older U.S. Military Veterans

	No/Minimal PTSD Symptoms N = 2,387 Weighted 88.5% 1 Mean (SE)	Subthreshold PTSD N = 262 Weighted 9.6% 2 Mean (SE)	Full PTSD N = 68 Weighted 1.9% 3 Mean (SE)	Test of Difference (F)	p	Pairwise Contrasts	Cohen d Subthreshold PTSD Versus No/Minimal PTSD Symptoms	Cohen d Full PTSD Versus No/Minimal PTSD Symptoms
Mental component summarya	53.6 (0.5)	49.4 (0.6)	45.3 (1.0)	72.62	<0.001	1>2>3	0.18	0.34
Vitality	49.3 (0.6)	47.7 (0.8)	47.2 (1.4)	4.52	0.011	1>2	0.06	−
Social functioning	48.2 (0.6)	44.5 (0.7)	41.0 (1.3)	33.29	<0.001	1>2>3	0.13	0.25
Role emotional	49.5 (0.4)	46.8 (0.5)	45.0 (0.9)	38.72	<0.001	1>2>3	0.14	0.23
Mental health	52.4 (0.4)	48.3 (0.5)	44.2 (0.9)	86.50	<0.001	1>2>3	0.22	0.42
Physical component summarya	43.7 (0.8)	40.6 (1.0)	40.0 (1.7)	10.80	<0.001	1>2,3	0.08	0.10
Physical functioning	44.4 (0.8)	41.6 (0.9)	40.9 (1.6)	10.73	<0.001	1>2,3	0.07	0.09
Role physical	45.3 (0.7)	42.5 (0.9)	40.8 (1.5)	13.28	<0.001	1>2,3	0.09	0.13
Bodily pain	46.3 (0.7)	42.3 (0.8)	41.2 (1.5)	24.75	<0.001	1>2,3	0.12	0.15
General health	47.4 (0.6)	45.1 (0.7)	43.7 (1.3)	10.70	<0.001	1>2,3	0.08	0.13
Cognitive functioningb	89.6 (0.9)	80.8 (1.1)	70.9 (1.9)	96.21	<0.001	1>2>3	0.21	0.43
Executive function	92.0 (1.2)	83.1 (1.4)	71.0 (2.5)	66.95	<0.001	1>2>3	0.16	0.36
Memory	82.8 (1.4)	73.7 (1.7)	66.3 (3.0)	36.80	<0.001	1>2>3	0.14	0.24
Attention	89.1 (1.3)	79.4 (1.5)	66.6 (2.7)	67.02	<0.001	1>2>3	0.16	0.36
Concentration/thinking	90.4 (1.1)	79.2 (1.3)	66.7 (2.4)	103.73	<0.001	1>2>3	0.22	0.45
Confusion	92.3 (1.0)	85.1 (1.2)	77.1 (2.1)	55.95	<0.001	1>2>3	0.15	0.31
Processing speed	91.1 (1.1)	84.4 (1.3)	77.9 (2.3)	38.58	<0.001	1>2>3	0.13	0.25
Psychosocial Difficultiesc	9.0 (0.9)	15.0 (1.0)	18.2 (1.8)	40.60	<0.001	3,2>1	0.14	0.21
Trouble with spouse/partner	0.7 (0.1)	0.9 (0.1)	1.2 (0.2)	6.97	0.006	2,3>1	0.04	0.10
Trouble with children	0.6 (0.1)	0.9 (0.1)	0.8 (0.2)	6.37	<0.001	2>1	0.06	–
Trouble with extended family relationships	0.5 (0.1)	0.9 (0.1)	0.8 (0.2)	9.56	<0.001	2>1	0.09	−
Trouble with friendships/socializing	0.5 (0.1)	1.0 (0.1)	1.2 (0.2)	18.76	<0.001	2,3>1	0.11	0.15
Trouble with daily activities	1.2 (0.1)	1.7 (0.1)	1.3 (0.2)	9.54	<0.001	2>1	0.11	−

Notes. SE: standard error of the mean.

Means are adjusted for age, gender, race/ethnicity, marital status, enlistment/commissioned status, combat veteran status, number of deployments, VA healthcare status, adverse childhood experiences, total lifetime trauma burden, and lifetime major depressive, alcohol, drug, and nicotine use disorders.

Degrees of freedom for ANCOVAs = 2, 1,989.

^aCurrent mental (α = 0.78) and physical health (α = 0.87)-related functioning were assessed using the Short-Form 8 Health Survey. Ware JE, Kosinski M, Dewey JE, et al: How to score and interpret single-item health status measures: A manual for users of the SF-8 Health Survey. Boston: QualityMetric Incorporated, 2001

^bCurrent cognitive functioning was assessed using the Medical Outcomes Study Cognitive Functioning Scale−Revised (α = 0.90), which assesses self-reported difficulties in six cognitive domains—executive function, memory, attention, concentration/thinking, confusion, and processing speed. Yarlas A, White MK, Bjorner JB. The development and validation of a revised version of the medical outcomes study cognitive functioning scale (MOS-COG-R). Value Health. 2013;16(3):A33–A34.

^cCurrent psychosocial difficulties were assessed using the Brief Inventory of Psychosocial Functioning (α = 0.74), which assesses functional difficulties in seven domains: romantic relationships, family relationships, work, friendships and socializing, parenting, education, and self-care. Kleiman SE, Bovin MJ, Black SK, et al. Psychometric properties of a brief measure of posttraumatic stress disorder-related impairment: The Brief Inventory of Psychosocial Functioning. Psychol Serv. May 2020;17(2):187–194. doi:10.1037/ser0000306.

Data Analysis

Raw, unweighted n’s are reported and all inferential analyses are weighted to permit generalizability to the U.S. veteran population. Data analyses proceeded in four steps. First, we computed independent-samples t tests and χ² tests to compare the three groups (i.e., full PTSD, subthreshold PTSD, and no or minimal PTSD symptoms) with respect to sociodemographic, military, and trauma characteristics. Second, we conducted a multinomial logistic regression analysis to identify sociodemographic, military, and trauma characteristics that independently differentiated veterans with subthreshold and full PTSD; characteristics that were associated with group status at the p < 0.05 level in bivariate analyses were entered into this model. Planned post-hoc analyses of trauma exposure variables were then conducted to identify specific exposures associated with subthreshold and full PTSD. Third, we conducted a series of binary logistic regression analyses to examine the relation between PTSD status and psychiatric and mental health treatment variables. Sociodemographic, military, and trauma characteristics that differed by PTSD status at the p < 0.05 level were adjusted for in these analyses, and analyses of suicidal behaviors and mental health treatment were additionally adjusted for lifetime major depressive, alcohol, drug, and nicotine use disorders. Fourth, we conducted one-way analyses of covariance to examine the relation between PTSD status and measures of functioning; composite scores were analyzed together in one model; and then subsequent models were conducted to examine mental and physical health subscales; and cognitive functioning domains. Sociodemographic, military, and trauma characteristics that differed by PTSD status at the p < 0.05 level and lifetime major depressive, alcohol, drug, and nicotine use disorders were adjusted for in these analyses. To balance risk of Type I and Type II error, and to explore the full range of health and functional measures associated with subthreshold and full PTSD, we employed a liberal alpha threshold of 0.05 for all analyses.

RESULTS

Prevalence of Subthreshold and Full PTSD

In the full sample of 3,001 older U.S. veterans, 262 (8.5%, 95% confidence interval [CI]: 7.4%–9.7%) screened positive for subthreshold PTSD and 68 (1.7%; 95% CI: 1.2%–2.3%) for full PTSD. Of the veterans who reported exposure to one or more potentially traumatic events (n = 2,821; 92.7%), the prevalence for subthreshold PTSD was 9.6% (95% CI: 8.4%–10.9%) and was 1.9% (95% CI: 1.3%–2.6%), for full PTSD. The prevalence of subthreshold (19.3%, 95% CI: 11.7%–29.1% versus 9.2%, 95% CI: 7.9%–10.5%) and full (6.8%, 95% CI: 2.5%–14.2% versus 1.7%. 95% CI: 1.2%–2.3%) PTSD were significantly higher in female relative to male veterans (χ²(2) = 23.12, p < 0.001). Further, the prevalence of subthreshold (16.6%, 95% CI: 12.9%–20.8% versus 8.2%, 95% CI: 6.9%–9.5%) and full (5.4%, 95% CI: 3.3%–8.3% versus 1.1%, 95% CI: 0.7%–1.8%) was significantly higher in veterans who reported using the VA as their primary source of healthcare relative to veterans who reported using another primary source of healthcare.

Sociodemographic, Military, and Trauma Characteristics

As shown in Table 1, the PTSD groups differed with respect to all of the sociodemographic and military variables assessed except for education, retirement status, household income, and duration of military service. With regard to trauma characteristics, the subthreshold and full PTSD groups endorsed more adverse childhood experiences and total traumas than the no/minimal PTSD symptom group. The prevalence of direct trauma exposures generally increased as a function of PTSD status.

TABLE 1.

Demographic, Military, and Trauma Characteristics by Current PTSD Status in Older U.S. Military Veterans

	No/Minimal PTSD Symptoms N = 2,387 Weighted 88.5% 1	Subthreshold PTSD N = 262 Weighted 9.6% 2	Full PTSD N = 68 Weighted 1.9% 3	Test of Difference (F or χ2)	p	Pairwise Contrasts
	Weighted Mean (SD) or n (Weighted %)	Weighted Mean (SD) or n (Weighted %)	Weighted Mean (SD) or n (Weighted %)
Age	73.2 (8.2)	70.8 (7.7)	69.4 (5.6)	10.40	<0.001	3,2>1
Male gender	2,248 (96.5%)	220 (91.6%)	52 (85.0%)	23.12	<0.001	1>2,3
Caucasian race/ethnicity	2,087 (86.5%)	215 (79.9%)	53 (80.0%)	7.67	0.022	1>2
College graduate or higher	1,153 (35.2%)	99 (27.1%)	26 (30.0%)	5.68	0.058	–
Married/partnered	1,739 (75.1%)	173 (63.5%)	43 (67.5%)	13.64	0.001	1>2
Retired	1,687 (70.1%)	179 (70.4%)	51 (78.0%)	1.21	0.55	–
Annual household income >$60k	1,372 (57.2%)	133 (52.2%)	32 (48.8%)	2.91	0.23	–
Enlisted/Commissioned into military	2,012 (81.7%)	230 (88.7%)	59 (82.5%)	6.14	0.047	2>1
Combat veteran	752 (29.7%)	101 (40.4%)	35 (58.5%)	24.13	<0.001	3>2>1
Number of deployments				29.49	<0.001
0	1,637 (70.8%)	160 (59.9%)	33 (42.5%)			3>2>1
1	499 (19.8%)	57 (23.3%)	21 (32.5%)			3>1
2+	234 (9.4%)	43 (16.8%)	14 (25.0%)			3,2>1
10+ years of military service	766 (30.3%)	93 (32.0%)	23 (39.0%)	1.61	0.45	–
VA is primary source of healthcare	363 (15.3%)	81 (29.9%)	32 (50.0%)	57.56	<0.001	3>2>1
Adverse childhood experiences	1.0 (1.5)	2.1 (2.1)	2.7 (2.6)	50.84	<0.001	3>2>1
Total traumas	7.9 (7.0)	12.1 (9.5)	12.7 (11.3)	34.28	<0.001	3,2>1
Direct traumas	3.0 (0.1)	4.8 (0.2)	5.7 (0.4)	79.15	<0.001	3>2>1
Indirect traumas	4.9 (0.1)	7.3 (0.4)	7.0 (1.1)	14.53	<0.001	2>1
Types of direct trauma exposures
Natural disaster (e.g., flood, hurricane, tornado, earthquake)	1,175 (48.8%)	148 (56.5%)	45 (70.0%)	10.86	0.004	3>1
Transportation accident (e.g., car accident, boat accident, train wreck, plane crash)	1,511 (66.5%)	188 (72.4%)	47 (68.4%)	33.78	<0.001	3>2>1
Physical assault (e.g., being attacked, hit, slapped, kicked, beaten up)	613 (26.5%)	130 (49.7%)	40 (63.2%)	2.82	0.24	–
Combat or exposure to a war-zone (in the military or as a civilian)	654 (26.4%)	104 (40.6%)	34 (56.1%)	8.20	0.017	–
Fire or explosion	387 (19.2%)	75 (27.1%)	33 (55.6%)	26.68	<0.001	3>2>1
Life-threatening illness or injury	672 (29.1%)	114 (46.4%)	37 (55.0%	66.99	<0.001	3,2>1
Assault with a weapon (e.g., being shot, stabbed, threatened with a knife, gun, bomb)	359 (16.2%)	74 (30.9%)	37 (52.6%)	55.63	<0.001	3>2>1
Exposure to toxic substance (e.g., dangerous chemicals, radiation)	446 (20.1%)	80 (28.8%)	31 (50.0%)	56.05	<0.001	3,2>1
Serious accident at work, home, or during recreational activity	512 (23.2%)	81 (28.9%)	27 (39.5%)	20.88	<0.001	3,2>1
Other unwanted or uncomfortable sexual experience	196 (8.0%)	52 (14.8%)	20 (24.3%)	33.94	<0.001	3,2>1
Sexual assault (rape, attempted rape, made to perform any type of sexual act through force or threat of harm)	79 (2.9%)	36 (10.0%)	16 (22.2%)	8.16	0.017	3>1
Serious injury, harm, or death you caused to someone else	86 (4.1%)	30 (11.8%)	15 (21.6%)	34.78	<0.001	3,2>1
Severe human suffering	55 (2.4%)	19 (7.7%)	9 (13.2%)	29.15	<0.001	3,2>1
Sudden violent death (e.g., homicide, suicide)	58 (3.0%)	16 (6.6%)	6 (10.3%)	12.24	0.002	3,2>1
Captivity (e.g., being kidnapped, abducted, held hostage, prisoner of war)	23 (1.0%)	6 (1.6%)	5 (6.3%)	12.04	0.002	2>1
Sudden accidental death	86 (3.9%)	19 (9.3%)	3 (5.3%)	41.85	<0.001	3,2>1
Any other very stressful event or experience	440 (17.9%)	117 (44.1%)	38 (56.8%)	101.00	<0.001	3,2>1
Duration of PTSD symptoms (years)	10.6 (15.1)	20.1 (19.2)	32.5 (19.4)	59.52	<0.001	3>2>1
PCL-5 score	2.7 (3.8)	22.4 (10.0)	43.6 (12.4)	2381.22	<0.001	3>2>1

Notes. Tests of difference were F tests from univariate analyses of variance (ANOVAs) for continuous variables and χ² for categorical variables; degrees of freedom for all ANOVAs = 2, 2,034; and for all χ² analyses = 2.

Index Traumas

The most prevalent index traumas in the full PTSD group were combat exposure (32.5%), physical assault (12.5%), sexual assault (7.5%), and life-threatening illness or injury (5.0%). The most prevalent index traumas in the subthreshold PTSD group were combat exposure (18.1%), life-threatening illness or injury (9.3%), exposure to sudden accidental death (8.3%), and physical assault (6.9%). Index traumas and categories of traumas did not differ significantly between veterans with full and subthreshold PTSD, all p’s < 0.05. However, in the full and subthreshold PTSD groups, female veterans were more likely than male veterans to endorse sexual assault as their index trauma (42.9% versus 1.1% in subthreshold PTSD group; 50.0% versus 2.9% in the full PTSD group).

Sociodemographic, Military, and Trauma Correlates of Subthreshold and Full PTSD

As shown in Table 2, female gender, unmarried/unpartnered status, VA as primary source of healthcare, and greater number of ACES and direct traumas were independently associated with subthreshold PTSD. Female gender, combat veteran status, VA as primary source of healthcare, and greater number of ACES and direct traumas were independently associated with full PTSD. Only a greater number of direct trauma exposures differentiated the full versus subthreshold PTSD groups.

TABLE 2.

Multinomial Regression Model of Demographic, Military, and Trauma Variables Associated With Subthreshold and Full PTSD in Older U.S. Military Veterans

	Subthreshold PTSD versus No/Minimal PTSD Symptoms OR (95% CI)	Full PTSD Versus No/Minimal PTSD Symptoms OR (95% CI)	Full PTSD Versus Subthreshold PTSD OR (95% CI)
Age	1.00 (0.98–1.02)	0.97 (0.92–1.03)	0.97 (0.92–1.04)
Female gender	1.94 (1.04–3.60)*	5.21 (1.76–15.40)**	2.69 (0.87–8.36)
Non-Caucasian race/ethnicity	1.04 (0.69−1.56)	0.72 (0.30–1.74)	0.69 (0.28–1.74)
Unmarried/unpartnered	1.55 (1.11–2.15)**	0.98 (0.47–2.06)	0.63 (0.29–1.37)
Drafted into military	0.74 (0.46–1.19)	1.52 (0.61–3.74)	2.05 (0.77–5.45)
Combat veteran	1.38 (0.99–1.92)	2.77 (1.30–5.88)**	2.00 (0.91–4.42)
VA is primary source of healthcare	1.62 (1.14–2.31)**	3.19 (1.62–6.33)**	1.97 (0.96–4.05)
Adverse childhood experiences	1.21 (1.15–1.32)***	1.30 (1.12–1.51)**	1.07 (0.92–1.25)
Direct traumas	1.23 (1.15–1.32)***	1.47 (1.27−1.69)***	1.19 (1.03–1.38)*
Indirect traumas	1.00 (0.98–1.02)	0.98 (0.94–1.02)	0.98 (0.94–1.02)

Notes. OR: odds ratio; 95% CI: 95% confidence interval.

Statistically significant association:

^*p < 0.05;

^**p < 0.01;

^***p < 0.001.

Degrees of freedom for each group comparison in multinomial regression model = 1.

Planned post-hoc analyses revealed that emotional neglect and having a family member with mental illness were the adverse childhood experiences independently associated with both subthreshold PTSD (OR: 2.97, 95% CI: 2.05–4.28 and OR: 1.82, 95% CI: 1.16–2.86, respectively) and full PTSD (OR: 2.74, 95% CI: 1.29–5.82 and OR: 3.53, 95% CI: 1.59–7.85), respectively.

Direct trauma exposures that were independently associated with subthreshold PTSD included combat exposure (OR: 2.27, 95% CI: 1.13–4.57) and physical assault (OR: 1.53, 95% CI: 1.01–2.30). Direct trauma exposures that were independently associated with full PTSD included fire/explosion (OR: 3.85, 95% CI: 1.59–9.29) and assault with a weapon (OR: 2.56, 95% CI: 1.01–6.44). The only direct trauma exposure that differentiated the full versus subthreshold PTSD group was fire/explosion (OR: 3.51, 95% CI: 1.39–8.86).

Co-occurring Psychiatric Disorders and Mental Health Treatment

As shown in Table 3, the prevalence of lifetime and current psychiatric disorders and mental health treatment increased as a function of PTSD status. In adjusted analyses, veterans with subthreshold PTSD were more likely than those with no/minimal PTSD symptoms to screen positive for lifetime MDD, AUD, DUD, NSSI, and SA; current MDD, GAD, AUD, DUD, and suicidal ideation; and to have ever and currently use mental health treatment. Veterans with full PTSD were more likely than those with no/minimal PTSD symptoms to screen positive for lifetime DUD, AUD, NSSI, and SA; current MDD, GAD, suicidal ideation and intent; and to have ever and currently use mental health treatment. Inspection of 95% CIs revealed that veterans with full PTSD were more likely than those with subthreshold PTSD to screen positive for lifetime MDD and current GAD, and to have ever received mental health treatment.

Functioning

Table 4 shows scores on measures of functioning by PTSD status, which decreased as a function of PTSD status for mental and cognitive functioning; and were reduced in both the subthreshold and full PTSD groups relative to the no/minimal PTSD symptom group for physical functioning and psychosocial difficulties. Effect sizes of group differences were small, with the most pronounced reductions for both the full and subthreshold PTSD groups on measures of concentration/thinking, overall cognitive functioning; and mental health-related functioning.

CONCLUSION

This study examined the prevalence, characteristics, and psychiatric and functional correlates of subthreshold and full PTSD in a nationally representative sample of older U.S. veterans. Approximately 10% of this population screened positive for current subthreshold PTSD and 2% for full PTSD. Rates were notably higher amongst veterans who use the VA as their primary source of healthcare, with nearly 17% reporting subthreshold PTSD (and 5% full PTSD), as well as among female veterans with 19% reporting subthreshold PTSD (and 7% full PTSD). These rates are consistent with, albeit somewhat smaller than existing estimates of PTSD in older veterans which range from 1% to 22% (pooled prevalence = 8%)²⁵ for PTSD and 7%–13%^4,17,32 for subthreshold PTSD.

These data point to the importance of further study and assessment of subthreshold PTSD in older veterans. Subthreshold PTSD and full PTSD shared common features including being more likely associated with index traumas of combat and physical assault. Other studies have similarly found that combat and assault are associated with higher odds of full PTSD − the present study extends the work to subthreshold PTSD. Subthreshold PTSD needs further study, including examination of relevant cut scores for subthreshold and full PTSD for older adults, which some previous work suggests should be lower than for younger adults.³³ In this study, older veterans who were unpartnered, experienced more adverse childhood experiences and more direct traumas, and who used the VA were at particularly elevated risk for subthreshold PTSD. The finding that emotional neglect and having a family member with mental illness were independently associated with subthreshold and full PTSD suggests that early difficulties with attachment may contribute to risk for clinically significant PTSD symptoms in older veterans. This finding is consistent with prior work documenting the adverse mental and physical health effects of adverse childhood experiences,³⁴ as well as a link between attachment insecurity and PTSD.^35,36 Taken together, these results underscore the role of early childhood adversity in the etiology of subthreshold and full PTSD in older U.S. veterans. They further indicate a high prevalence and clinical burden of subthreshold PTSD in addition to full PTSD in this population.

These findings add to the relatively sparse data on PTSD in older women³⁷ and suggest a high burden of subthreshold and full PTSD in older female veterans. In a large nationally representative sample of U.S. adults, one of seven older women reported physical or sexual assault or both, and those who did were more likely than those without such histories to meet criteria for past-year and lifetime PTSD, depression, or anxiety.³⁸ Furthermore, a recent study using national data from VA medical records from over 70,000 women aged 55 and older found that over 13% had a positive screen for military sexual trauma and those that did had higher rates of PTSD, depression, and suicidal ideation, as well as multiple medical conditions, particularly sleep disorders and pain.³⁹ Relatedly, in a study of over 8,000 women who served during the Vietnam era, the prevalence of PTSD was higher than previously documented and was associated with several wartime exposures, not only exposure to death and dying.⁴⁰ Taken together, these findings suggest that trauma exposure, as well as subthreshold and full PTSD, are prevalent and strongly linked to mental health outcomes in both older civilian and female veterans. They further highlight the importance of greater outreach, screening and treatment of subthreshold and full PTSD in older female veterans. The issue of PTSD in older female veterans will be important to continue to follow, especially given that increasing numbers of women are serving in the military.

Subthreshold PTSD also carries similar risks as does full PTSD in its association with suicide attempt and nonsuicidal self-injury. Given that older veterans aged 55–74 are at the highest risk for dying by suicide,⁴¹ better recognition of both subthreshold and full PTSD may be an important component of suicide prevention efforts. Relatedly, it is noteworthy that subthreshold PTSD carries similar risks to full PTSD in its association with reduced function and mental health comorbidity. As both physical disability,⁴² cognitive changes,⁴³ and depression⁴⁴ are associated with suicide risk in older veterans, these represent important aspects of comorbidity and may also be associated with heightened risk for suicide. These findings suggest the importance of assessing for and identifying older veterans with subthreshold PTSD as part of population-based suicide prevention efforts that consider the intersection of social and trauma risk factors in suicide prevention. It is noteworthy that even though subthreshold PTSD is not a widely recognized diagnostic entity, older veterans who screened positive for it were three times more likely than those with no/minimal PTSD symptoms to be engaged in mental health treatment. One interpretation of this finding is that treatment targeting subthreshold symptoms may prevent the development of more clinically severe PTSD symptoms and decrease risk for comorbid psychiatric and functional problems. Another possibility is that with treatment, some older veterans who previously met criteria for full PTSD experienced a decrease in symptoms such that they endorsed a level of symptomatology consistent with subthreshold PTSD at the time of the survey.

Older adults may re-engage with rather than avoid memories of trauma²¹ as they confront illness, bereavement, role transitions, and cognitive changes and seek to form meaning and build cohesion in later life.²² As such, older adults with full or subthreshold PTSD may present with more engagement with traumatic memories and less avoidance,²¹ which is important to understand when assessing and providing treatment. Given lower rates of PTSD in older versus younger adults,⁴⁵ many older adults may confront memories of earlier trauma with psychological resilience. Further, older veterans who are re-engaging with traumatic memories as part of life review, may not meet criteria for full PTSD if they are not actively avoiding thoughts and external reminders of a traumatic event. Nevertheless, given the concerning clinical correlates of both subthreshold and full PTSD in older adults, including suicidality, comorbid disorders, and functional difficulties, it remains important to carefully identify PTSD in and provide treatment for older adults in general, and older veterans in particular. Attention to veteran status is particularly important regardless of setting as in the U.S., most veterans do not receive care within the VA healthcare system.²³

Limitations of this study must be noted. First, we used self-report measures to assess PTSD and other conditions, which could minimize or inflate symptomatology. Second, we assessed a broad range of health and functional outcomes and employed a liberal alpha threshold set to 0.05; further research using larger samples would be helpful in evaluating the generalizability of the results reported herein. Third, the cross-sectional study design does not allow us to disentangle longitudinal or causal relationships between PTSD status and clinical correlates. Fourth, while the NHRVS sample is nationally representative of U.S military veterans, results may not generalize to non-veteran populations, which are more demographically diverse. Potentially differing prevalence rates based on criteria used should be considered in evaluating study outcomes.⁴⁶

Notwithstanding these limitations, results of this study suggest that subthreshold and full PTSD are prevalent in older U.S. veterans, especially among female veterans and veterans who use the VA as their primary source of healthcare. Subthreshold PTSD is associated with a comparable clinical and functional burden as full PTSD, thus underscoring the importance of assessing, monitoring, and treating both of these manifestations of PTSD symptoms in clinical settings.

Highlights.

• This study provides estimates of PTSD and associated comorbidities in a nationally representative sample of older U.S. military veterans.

• Overall, 9.6% of veterans had subthreshold PTSD and 1.9% full PTSD; but was higher in women and those who use VA as their main source healthcare. Veterans with subthreshold PTSD were equally as likely as those with full PTSD to have psychiatric, mental, cognitive, and functional comorbidities, including a history of suicide attempts and current suicidal ideation.

• Clinicians should attend to subthreshold and full PTSD in older adults.